Urban Ecosyst (2006) 9:243–257DOI 10.1007/s11252-006-9354-2

Effects of urbanization on plant flowering phenology:A review

Kaesha Neil · Jianguo Wu

Published online: 14 July 2006C© Springer Science + Business Media, LLC 2006

Abstract Studies of flowering and leafing phenology have dramatically increased duringthe last few decades because changes in plant phenology can be indicative of possible effectsof climate change at multiple scales. This article reviews the available literature focusing onthe effects of urbanization on flowering phenology. The literature of flowering phenologyin urban environments suggests that spring-blooming plants in a variety of ecosystems inNorth America, Europe, and China tend to bloom earlier in the city than in the surroundingun-urbanized habitat. Moreover, ephemerals, early spring bloomers, and insect-pollinatedplants in these environments tend to be more sensitive than perennials, mid- or late-springbloomers, and wind-pollinated plants. Researchers attribute advanced flowering in urbanenvironments to the Heat Island Effect. The potential ecological consequences of changesin flowering phenology in urbanized areas are not well understood or explicitly studied.However, studies in global biology have suggested that climate change may result in aseries of important ecological consequences as well as human-related problems such asearlier and extended allergy seasons. More field-based studies are needed to elucidate thisissue.

Keywords Urban ecology . Flowering phenology . Urban heat island . Ecosystem structure

Introduction

Urban climate conditions are considered similar to the changing global climate conditions;therefore, many researchers study urbanized areas as small-scale experiments, or models, ofglobal climate change (Ziska et al., 2003). Studies of plant phenology, the seasonal timing ofenvironment-mediated growth and reproduction events in plants and animals, and changesin phenology caused by urbanization will improve our understanding of intra- and inter-specific plant interactions and potential changes in their interactions in novel environments

K. Neil (�)· J. WuLandscape Ecology and Modeling Laboratory (LEML), School of Life Sciences, Arizona StateUniversity, P.O. Box 874601 Tempe, AZ 85287-4601, USAe-mail: kaesha.neil@asu.edu

Springer

244 Urban Ecosyst (2006) 9:243–257

(Rathcke and Lacey, 1985; Mussey and Potter, 1997; Bishop and Schemske, 1998; Penuelasand Filella, 2001; Petit, 2001). Through studying these interactions in urban ecosystems, anopportunity is presented to test ecological theories (Pickett et al., 2001; Wu and David, 2002)related to coevolution, selective forces, community dynamics, population dynamics, andphenotypic plasticity. Indeed, the recent surge in phenological studies has occurred becausephenology is a highly sensitive indicator that can be used to study the possible effects ofclimate change at multiple scales (Chuine et al., 2000; Schwartz and Reiter, 2000; Sparks andMenzel, 2002; Zhao and Schwartz, 2003; Penuelas et al., 2004; Williams and Abberton, 2004).Following phenological events also allows researchers to analyze net primary productivity(Schwartz et al., 2002; Badeck et al., 2004), energy and mass balance in the lower atmosphere(Schwartz and Chen, 2002; Kang et al., 2003), and multiscale plant dynamics (Schwartz andChen, 2002).

Direct observations, remote sensing (via satellite), and modeling tools and techniqueshave been used to study plant phenology at regional scales. For example, onset of spring inWisconsin (Zhao and Schwartz, 2003), China (Schwartz and Chen, 2002), the mid-latitudes ofNorth America (Schwartz and Reiter, 2000), and Korea (Kang et al., 2003) have been studiedusing a combination of observation records, remote sensing, and modeling. However, Badecket al. (2004) warns that direct ground observations and remote sensing data do not measure thesame variables and researchers should consider this when they are using both data collectionmethods. These long-term studies indicate earlier green-up (start of seasonal photosynthesis)in high and mid latitude areas across the globe, although this is not a homogenous response(Penuelas and Filella, 2001; Sparks and Menzel, 2002; Walther et al., 2002; Schwartz, 2003;Badeck et al., 2004).

Selective forces acting on the timing of plant phenological events include abiotic and bioticelements, alone or in combination. Availability of light, nutrients, temperature, and moistureare key abiotic factors that may directly or indirectly limit the reproductive season of plants(Rathcke and Lacey, 1985). Biotic factors such as pollinators, pathogens, and herbivoresare also recognized as potential selective forces on timing of flowering (Rathcke and Lacey,1985; Brody, 1997).

Flowering phenology is often quantified by determining the time of onset and end offlowering, duration of flowering, skewness, and number of flowers produced (Augspurger,1983; Rathcke and Lacey, 1985). Gene-environment interactions determine the timing ofphenological events (e.g., new leaves of deciduous plants, flowers, and fruit). Events thatbreak dormancy and initiate growth and reproductive development are called phenologicaltriggers (Bowers and Dimmitt, 1994). Photoperiod, temperature, and soil moisture have beenrecognized as the main environmental triggers for leafing and flowering (Rathcke and Lacey,1985); although many plants rely on multiple cues that may occur months apart or respondto different cues depending on the season (Yang and Abe, 1974; Rathcke and Lacey, 1985;Friedel et al., 1993; Bowers and Dimmitt, 1994). For example, some temperate forest plantsrequire a chilling cue to begin biochemical reactions that will allow them to flower in responseto the photoperiod cue several months later (Rathcke and Lacey, 1985). Bowers and Dimmitt(1994) found that several Sonora Desert perennials need both moisture and temperature cuesto flower.

Photoperiod and temperature are regarded as the primary triggers of phenological eventsin many temperate forest, boreal, and Mediterranean plants. Unfortunately, temperature isoften stated as the flowering trigger even though it is the interaction between temperatureand photoperiod that is responsible for initiating flowering (Zhang et al., 2004). Moisture isthought to be the primary trigger for arid and semi-arid plants, but empirical studies have

Springer

Urban Ecosyst (2006) 9:243–257 245

shown that temperature and photoperiod also are important (Friedel et al., 1993; Bowers andDimmitt, 1994; Abd El-Ghani, 1997).

The purpose of this paper is to synthesize the known spatiotemporal patterns of floweringin cities, the effects of urbanization on flowering phenology and potential consequences,and address unanswered questions and areas of research requiring further study, such asthe mutual affects of landcover patterns of cities, spatiotemporal pattern of flowering withincities, ecological community structure and function within cities, and spatiotemporal patternof flowering affect on human health and agriculture.

Patterns of flowering phenology in response to triggers

While there have been a number of recent studies of flowering and leafing phenology, themajority have focused on areas outside of urban ecosystems (Penuelas and Filella, 2001;Sparks and Menzel, 2002; Walther et al., 2002; Badeck et al., 2004). Nevertheless, fieldobservations and remote sensing studies conducted across a range of habitats and in geo-graphically widespread areas show an advancement of flowering phenology and greening(remote sensing used for the latter case only) anywhere from a few days to a couple of weeksin urban areas compared to the surrounding rural environment (Roetzer et al., 2000; Fitterand Fitter, 2002; White et al., 2002; Ziska et al., 2003; Zhang et al., 2004). These results areseen for both ephemerals and perennials; however, not all plants respond the same way. Earlyspring bloomers show more advancement than mid- or late-spring bloomers (Fitter and Fitter,2002). In addition, animal-pollinated plants show more advancement than wind-pollinatedplants (Fitter and Fitter, 2002). It is not clear from the current literature whether the samepattern exists in urban areas located in different biomes or ecosystem regions. There is alsoa paucity of information regarding the spatiotemporal pattern within cities. Existing studiesonly differentiate between urban and rural areas without looking at finer spatial or temporal(intra-annual) scales.

Historical temporal patterns

Phenological studies in urban areas of England, several central European countries, andBoston, Massachusetts, USA, demonstrate a historical trend of advanced timing of flow-ering over the last 50–100 years (Roetzer et al., 2000; Fitter and Fitter, 2002; Hepper,2003; Primack et al., 2004). However, Hepper (2003) also reported that many gardenplants at the Royal Botanical Gardens, England, demonstrated a delay or no change inflowering. The Royal Botanical Gardens is located within a large metropolitan area, about10 km outside of London, near Richmond. Fitter and Fitter (2002) noted high variabil-ity in plant flowering time changes. For instance, many of their 385 plants did not showa significant change while only 3% of the 24% of the plants that flowered later weresignificant.

Both Fitter and Fitter (2002) and Roetzer et al. (2000) observed greater advances in theflowering time of some plants post 1980s and 1990s compared to earlier in the century.Roetzer et al. (2000) also observed a lack of difference in responses between urban andrural areas over time. Roetzer et al. (2000) stated the high variability may be a result ofhigher temperature trends in early spring and lower temperature trends in late spring as wellas great spatial variability in flowering responses. A key issue is how much of the histori-cal changes are attributable to global warming versus temperature changes associated withurbanization.

Springer

246 Urban Ecosyst (2006) 9:243–257

Spatial patterns

A clear difference in timing of flowering between urban and rural areas has been observedin the United States, Europe, and Asia (Roetzer et al., 2000; White et al., 2002; Ziskaet al., 2003; Zhang et al., 2004). Zhang et al. (2004) used the moderate-resolution imagingspectroradiometer (MODIS) to study the timing of green-up and dormancy in the northernhemisphere (between 35◦N and 70◦N) in 2001. Their data show vegetation in large urbanareas greening about 4–9 days earlier and going dormant about 2–16 days later than in thesurrounding rural environment. This could imply a similar large-grain response in flowering.Using Advanced Very High Resolution Radiometer (AVHRR) data from 1990s, White et al.(2002) studied phenological differences in plant green-up between urban and rural areas ineastern United States broadleaf forests. They observed an expansion of the growing seasonof about 7.6 days in urban areas. Most of the expansion was due to an earlier start of thegrowing season and the rest was due to delayed dormancy. Not all areas showed the sametrends; for instance, some areas showed a delay in green-up.

Ziska et al. (2003) studied ragweed (Ambrosia artemisiifolia) in Baltimore, Maryland,USA along an urban-rural gradient. Ragweed grew faster, flowered earlier, senesced earlier,and had greater above-ground biomass at the urban sites than at the rural sites, althoughthe pollen allergenicity was greater at the rural sites. Flowering time of ragweed occurredearlier in the season the closer the plants were to the urban center. Roetzer et al. (2000)analyzed data for four spring blooming plants from 10 central European observation sta-tions between 1951 and 1995. At almost all stations (except Zurich), the plants tended tobloom earlier in urban areas than in surrounding rural areas by about 2–4 days. However,not all differences were significant and there was great variability, part of which mightbe a result of observation stations not being in completely urbanized or completely ruralareas.

Even though most of the above-mentioned studies demonstrate a difference between ruraland urban areas, finer-scale spatial resolution patterns within cities are not completely clear.Will the same patterns be observed among cities? Also, the phenological studies indicatevariability in flowering responses. Do different land cover/land use types affect floweringphenology differentially? Will this result in a multiscale pattern from local (neighborhood)to regional (whole cities or groups of cities) scales? For example, will there be differencesbetween urban commercial, residential, xeriscaping, and mesicaping? This is particularlyinteresting when one considers cities not built in a concentric pattern. Will the urban-ruralgradient pattern observed by Ziska et al. (2003) be observed in multi-nuclei, or polycentricand patchy, cities like Phoenix in which agriculture, residential, and commercial land coverare interspersed (Luck and Wu, 2002)?

Patterns by plant functional type

Several of the studies reviewed reported differences in plant responses based on plantfunctional type. Insect-pollinated plants tended to have greater advancement of floweringthan wind-pollinated plants (Fitter and Fitter, 2002), annuals demonstrated greater advance-ment than perennials (Fitter and Fitter, 2002), and early spring bloomers showed greateradvancement than mid- and late-spring bloomers (Fitter and Fitter, 2002; Hepper, 2003;Traidl-Hoffmann et al., 2003). Moreover, Zhang et al. (2004) observed that forest land coveris more sensitive (showed more advancement of green-up) than savanna land cover, whichshowed more advancement than grass land cover. United States urban areas tend to showmore advancement than European or Asian urban areas. Thus, most studies suggest that

Springer

Urban Ecosyst (2006) 9:243–257 247

short-lived plants (with life history strategies that avoid environmental stresses) and plantsreliant upon animals for reproductive success may be affected the most by urbanization-caused environmental change. This makes sense because these types of plants are known tobe highly responsive to environmental variables. However, questions remain: Do self-fertileplants differ in response from obligate out-crossing plants? Do generalist pollinator plantsdiffer in response from specialist pollinator plants? Are there general characteristics that de-fine the types of plants blooming later vs. earlier? How are plants that flower in the summer,fall, or winter affected by urban climate conditions?

Causes of flowering phenological changes in urban environments

A mechanistic understanding of the causes and processes of changing flowering phenology inurban environments is severely lacking (Battey, 2000) although there have been an increasingnumber of studies on flowering phenology in cities (Table 1). In the following section, wesummarize the major factors believed to cause changes in plant flowering phenology.

Temperature

All the studies of flowering phenology in cities attributed the advancement in floweringto the urban heat island effect. Zhang et al. (2004) stated different land cover types intheir study demonstrated differential responses to temperature as a phenological trigger:forest land cover type being the most sensitive and grassland cover type the least sen-sitive. Zhang et al. (2004) also attributed the difference in advancement between UnitedStates urban areas and European and Asian urban areas to higher urban temperatures: 1–3◦Cwarmer in US cities compared to surrounding rural environments and about 0.8◦C warmerin European and Asian cities compared to surrounding rural environments (Zhang et al.,2004). White et al. (2002) also hypothesized the delay in green-up in some regions was aresult of higher winter temperatures, which might have delayed the fulfillment of chillingrequirements.

Fitter and Fitter (2002) suggested that insect-pollinated plants and annuals are moresensitive to temperature than wind-pollinated plants or perennials as a result of life his-tory strategies that depend heavily on temperature as a cue for phenological events. Ziskaet al. (2003) reported a high correlation between temperature/CO2 concentration and earlierflowering time, but did not separate temperature effects from CO2 effects. Hepper (2003)attributed observed advancement in flowering to global warming, but urbanization is morelikely to have played a role at the local scale since local landscape characteristics (e.g., patternof built features, vegetation amount, etc.) strongly affect microclimatic conditions. Primacket al. (2004) reported a correlation between winter and spring temperatures and floweringphenology at the Arnold Arboretum of Boston, MA, USA. However, when they statisticallycontrolled for temperature in multiple regression analysis, they still found earlier flower-ing suggesting an additional cause. Other variables Primack et al. (2004) hypothesized ascauses for earlier flowering included changes in rainfall, humidity, land cover change at thearboretum affecting temperature (temperature was obtained from a weather station), and ageand size of the plants. Because they followed specific plants over the years, it was possiblethat increases in size and aging could have affected flowering time. This would need to beaddressed on a species by species basis.

It appears that increase in temperature is considered the primary cause of flower-ing advancement as global climate change studies and numerous laboratory experiments

Springer

248 Urban Ecosyst (2006) 9:243–257Ta

ble

1C

ited

liter

atur

eex

plic

itly

addr

essi

ngflo

wer

ing

phen

olog

yan

dgr

een-

up(l

eafin

g)in

urba

nec

osys

tem

s

Ref

eren

ceSt

udie

dL

ocat

ion

Peri

odFi

ndin

gs

Bla

nken

ship

etal

.,19

93E

aste

rlil

ies

and

the

effe

cts

oflo

ng-t

erm

expo

sure

toet

hyle

ne.

Gre

enho

use

inN

orth

Car

olin

a19

90–1

991

Tota

lbud

initi

atio

nde

crea

sed,

bud

abor

tion

incr

ease

d,m

any

flow

ers

did

noto

pen

com

plet

ely,

buds

abno

rmal

lycu

rved

,flo

wer

edea

rlie

r,an

dse

nesc

edea

rlie

r.R

oetz

eret

al.,

2000

Flow

erin

gtim

eof

four

pere

nnia

lsE

urop

e19

51–1

995

Adv

ance

dflo

wer

ing

insp

ring

inur

ban

area

sco

mpa

red

toth

esu

rrou

ndin

gru

rala

reas

by2–

4da

ys.E

arly

spri

ngbl

oom

ers

wer

em

ore

vari

able

butt

ende

dto

bem

ore

sens

itive

toth

ehe

atis

land

effe

ctth

anm

id-

tola

te-s

prin

gbl

oom

ers.

Fitte

ran

dFi

tter

2002

Firs

tflow

erin

gda

teof

ase

tof

557

plan

tsp

ecie

sA

sing

lelo

catio

nin

Sout

h-ce

ntra

lE

ngla

nd

1954

–200

038

5sp

ecie

ssh

owan

adva

ncem

ento

fflo

wer

ing

(inc

reas

ing

mor

ein

the

1990

s)an

d10

spec

ies

flow

ered

late

r.H

igh

vari

abili

ty.

Whi

teet

al.,

2002

Gre

en-u

pof

urba

nde

cidu

ous

broa

dlea

ffo

rest

sus

ing

AV

HR

R.

Eas

tern

Uni

ted

Stat

es19

90–1

999

Exp

ansi

onof

the

grow

ing

seas

onby

∼7.8

days

—m

osto

fit

from

anea

rlie

rst

art.

Zis

kaet

al.,

2003

The

effe

cts

ofur

bani

zatio

non

Am

bros

iaar

tem

isii

foli

aal

ong

anur

ban-

rura

lgr

adie

nt

Bal

timor

e,M

aryl

and,

USA

2000

–200

1H

ighe

rC

O2

conc

entr

atio

ns/te

mpe

ratu

res

alon

gan

urba

n-ru

ral

grad

ient

appe

arto

sign

ifica

ntly

affe

ctph

enol

ogy

(adv

ance

men

t),

alle

rgen

icity

(dec

reas

e),a

ndpo

llen

prod

uctio

n(i

ncre

ase)

.T

raid

l-H

offm

anet

al.,

2003

Rev

iew

edhe

alth

liter

atur

eon

polle

nosi

s.E

urop

e,N

orth

Am

eric

a–

Hig

hsp

atia

lvar

iabi

lity

inre

spon

ses

ofpl

ants

tocl

imat

ech

ange

such

as:p

rem

atur

est

arto

fth

epo

llen

seas

on,e

arlie

rpe

aks,

exte

nsio

nof

polle

npe

riod

for

som

ela

teflo

wer

ing

spec

ies,

decr

ease

innu

mbe

rof

days

with

polle

nco

unt,

and

incr

ease

dst

reng

thof

the

polle

nse

ason

.C

ape

2003

Rev

iew

edst

udie

sad

dres

sing

the

effe

cts

ofV

OC

son

plan

ts.

––

Mor

elo

ngte

rm,l

owdo

sest

udie

s(fi

eld

and

lab)

need

tobe

done

sinc

ecu

rren

tres

earc

his

inco

mpl

ete.

Hep

per

2003

Firs

tflow

erin

gda

tes

ofco

mm

onpe

renn

ial

gard

enpl

ants

Lee

dsan

dR

ichm

ond,

Bri

tain

1946

–200

2Sp

ring

plan

tste

ndto

bloo

mea

rlie

ran

dsu

mm

erpl

ants

tend

tobl

oom

late

r.Z

hang

etal

.,20

04L

eafin

gph

enol

ogy

(i.e

.,gr

een-

upan

ddo

rman

cy)

usin

gM

OD

ISM

idto

high

latit

udes

inth

eU

nite

dSt

ates

,E

urop

e,A

sia

2001

Gre

en-u

poc

curr

ed4–

9da

ysea

rlie

ran

ddo

rman

cy2–

16da

ysla

ter

inur

ban

area

sth

anin

adja

cent

natu

rala

reas

.Sug

gest

edU

HI

corr

elat

ion.

Prim

ack

etal

.,20

04Pe

akflo

wer

ing

date

sof

indi

geno

usan

dno

n-in

dige

nous

woo

dypl

ants

Arn

old

Arb

oret

um,

Bos

ton,

Mas

sach

uset

ts

1885

–200

3Pl

ants

show

edad

vanc

emen

tin

flow

erin

g(p

eak)

ofab

out8

days

;flo

wer

ing

in20

03(w

ithw

inte

rte

mpe

ratu

rew

ellb

elow

rece

ntno

rmal

tem

ps.)

show

edsi

mila

rtim

ing

asse

enin

earl

y19

00s.

Springer

Urban Ecosyst (2006) 9:243–257 249

demonstrate earlier flowering at higher temperatures. For example, Penuelas and Filella(2001) showed that the advancements of phenological events were significantly correlatedwith temperature increase over the last 30 years. However, correlation cannot be equated tocause-effect relationship. Moisture, CO2, volatile organic compounds (VOCs), photoperiod,humidity, and ultra-violet radiation may also cause changes in flowering phenology, althoughtheir importance is still controversial. Nevertheless, with accelerating urbanization expectedin the future, it is likely that increasing temperature due to urban heat islands will continue tobe an important cause for phenological changes. How much will urban temperatures changein the future? Will global warming magnify urban climate changes or mask them? Anotherinteresting question yet to be explored is: Will plants continue to advance flowering time?Or will a threshold temperature be reached?

Moisture and humidity

Penuelas et al. (2004) pointed out that many studies have ignored the potential role of soilmoisture on flowering phenology even though rainfall is known to affect phenology in sea-sonal tropical forests, in dry regions, and even in some places with “good water supply” likeNorway. While not conducted specifically in an urban area, the experimental field studies onplants in the Mediterranean conducted by Penuelas et al. (2004) demonstrated a correlationbetween decreased moisture availability and delayed flowering. Spatial variation in precip-itation across the landscape also can produce patchy flowering phenology in tropical areas(Rathcke and Lacey, 1985).

Humidity may affect flowering phenology of some plants, particularly as a secondarytrigger (Pavon and Briones, 2001), although the mechanism is unclear. Primack et al. (2004)pointed out that soil moisture or humidity could have contributed to earlier flowering intheir study. Urbanization can change humidity in an area either by increasing or decreasingsurface water and/or number of plants (Adebayo, 1991; Lipfert et al., 1991; Chow, 1992;Unger, 1999; Jonsson, 2004). How strong is the effect of humidity on flowering phenologyin urban areas as compared to temperature? Which plants are most sensitive to humiditychanges? These are questions to be addressed in future studies.

Photoperiod

While photoperiod and photoperiod-temperature interactions are known to affect floweringphenology (Deen et al., 1998; Yan and Wallace, 1998; Borchert and Rivera, 2001; Riveraand Borchert, 2001; Franklin and Whitelam, 2004), no studies have been found that haveexamined how urbanization might modify day length (night length) perception by plants.Light pollution (from buildings and outdoor lighting) has been studied, but only in rela-tion to impacts on animal life (e.g., Longcore and Rich, 2004), human health (e.g., Pauley,2004), energy conservation (e.g., Osman et al., 2001), and stargazing (e.g., Crawford, 2001).Changes in the far red/red light ratio would be most likely to cause problems since thoseare the wavelengths plants use for photoperiod perception. Moreover, experimental studiesdemonstrate that circadian clocks and light perception interact to trigger flowering. If thecircadian clock-light perception synchrony is disrupted by light (including artificial) at thewrong time, day/night perception may be disrupted (Yanovsky and Kay, 2003), which canretard growth and reproduction. Experimental (in situ and ex situ) studies are necessary todetermine if and how ecological light pollution affects flowering phenology and plant life ingeneral.

Springer

250 Urban Ecosyst (2006) 9:243–257

Other factors

Rusterholz and Erhardt (1998) studied the effects of increased CO2 concentration on twolegumes and two non-legumes in the context of predictions from global climate change modelsrather than changes caused by urbanization. The authors observed varied responses: no effect,advanced flowering, delayed flowering, and an increase in flower number. However, there iscontroversy over how higher CO2 concentrations influence flowering phenology and if it isspecies-specific (Badeck et al., 2004): directly via affecting internal mechanisms or indirectlyvia increased vegetation mass, which produces more flowers. Higher CO2 concentrations arealso found in urban areas relative to non-urbanized areas (Idso et al., 2001; Day et al., 2002;Nasrallah et al., 2003) and may affect the flowering phenology of some plants.

Cape (2003) hypothesized that volatile organic compounds (VOCs) also affect floweringphenology. VOCs do not appear to affect flowering phenology; however, Cape (2003) coun-ters that the studies usually only tested VOCs at high concentrations for less than two weeks.One of the exceptions was a study conducted with ethylene for a longer time period andwith lower concentrations in which floret reduction was observed. Another experiment con-ducted for more than 100 days with Easter lilies demonstrated decreased total bud initiation,increased bud abortion, incomplete flower opening, abnormally curved buds, earlier flower,and earlier senescence at ethylene concentrations (≥ 0.5 microliters per liter of air) similarto those found in urban areas (Blankenship et al., 1993). Since exposure to long periods(relative to plant life) of low amounts of VOCs is more realistic than short durations of highconcentrations for plants near roadsides, further experimental study is important. Chauhanet al. (2004) reported delayed flowering of Cassia siamea near heavily used roadsides com-pared to plants in a botanical garden. They attributed the difference to pollution given off byvehicles.

Another controversial potential influence on flower phenology is ultraviolet-B (UV-B)radiation. Concerns over ozone depletion and the resulting increase in UV radiation havespurred studies of the effects of UV radiation on plants. Studies on the effects of increasedUV-B show increased flowering duration (Conner and Neumeier, 2002), delayed floweringtime (Sampson and Cane, 1999), and delayed or no change in flowering phenology of cropplants (Kakani et al., 2003). In contrast to these studies is the effect of decreased UV-Bexposure in urban areas due to pollution (e.g., Papayannis et al., 1998; Repapis et al., 1998).If UV-B radiation affects flowering phenology, how does decreased UV-B exposure affectflowering? Which kinds of plants have flowering phenologies affected by UV-B? Is reductionin UV-B exposure caused by pollution strong enough to affect flowering phenology? Do otherurban environmental factors counteract any affects of UV-B? These questions still awaitanswers.

Consequences of changes in flowering phenology

Most, if not all, existing studies of urbanization effects on flowering phenology are descrip-tive, and associated hypotheses have largely not been directly tested. Many of the potentialconsequences of phenological changes discussed below are based on studies of global climatechanges. Their relevance here rests on the assumption that urbanization and global climatechange are similar in ways of affecting plant phenology through increasing temperatureaccompanied by elevated CO2. In the following, we discuss the consequences of pheno-logical changes in terms of ecological patterns and processes, human health, and economicramifications.

Springer

Urban Ecosyst (2006) 9:243–257 251

Ecological consequences

Changes in flowering phenology across an urban landscape have the potential to affect plantpopulation dynamics. Early and late flowering have been correlated with decreased seedset (Rathcke and Lacey, 1985; Santandreu and Lloret, 1999). In addition, spatial isolationdue to habitat fragmentation has been addressed by many researchers (e.g., Fischer et al.,2000; Aizen et al., 2002; Murren, 2002); however, temporal isolation has not. Even if plantswere not isolated spatially, if plants across an urban landscape flower at times that do notoverlap or have only a small amount of overlap then they can become reproductively isolated(Fitter and Fitter, 2002; Hendry and Day, 2005). Genetic variability may then be positively(e.g., keeping landscape and indigenous plants from interbreeding and local adaptation)or negatively (e.g., inbreeding depression) affected. This may not be a problem with non-indigenous plants that human entities do not want proliferating, but can be one for other non-indigenous plants or indigenous plants in which successful reproduction is desirable. On theother hand, since changes in flowering phenology are not the same for every plant, increasedoverlap in flowering time is also possible (Fitter and Fitter, 2002). Increased synchrony offlowering between different cultivars, subspecies, and species of the same genus may resultin increased hybridization and, therefore, genetic variability. Again, this may be desirable ornot because it may lead to out-breeding depression or prevent in-breeding depression.

The effects of urbanization on flowering phenology may become important at the com-munity level. Affects on other plant species, pollinators, herbivores, secondary consumers,and pathogens due to changes in flowering phenology must be considered. Depending on thespecies involved, a change in flowering time and/or duration may create an asynchronousmismatch between pollinators and their flowers. This is more likely for plants and animalsthat have different environmental cues for development. For example, Kudo et al. (2004)observed that some ephemerals had a smaller seed set than others during a couple of un-seasonably warm years. They hypothesized that the differences were caused by the type ofpollinator. The ephemerals with lower seed sets tended to be bee pollinated whereas theephemerals with relatively steady seed set tended to be pollinated by flies. Although all theephemerals tended to flower earlier, bees may not have come out of hibernation earlier andthe flies did not hibernate at all.

On the other hand, there is no guarantee that specialization will make a plant or its de-pendent pollinators more vulnerable. Wcislo and Cane (1996) found that even specializedbees will switch plants when their “normal” pollen source is unavailable—this benefits thebees and may have varied consequences for the plants they “normally” pollinate. More-over, although pollinators and plants are considered important mutualists, the importance ofspecific species interactions on long-term survival is still controversial (e.g., Rathcke andLacey, 1985). Are mutualistic relationships the result of opportunistic exploitation or strongco-evolutionary interactions? How we answer this question will influence how we interpretany potential changes in indigenous and non-indigenous animal-plant interactions.

Another example of community dynamics impacts is the European winter moth (Operoph-tera brumata), great tit nestlings (Parus major), and oak trees. O. brumata is not negativelyaffected by the earlier bud burst of the oak they feed on because they keep synchronous byhatching earlier. However, P. major, which eat O. brumata, do not hatch earlier (Waltheret al., 2002). This results in the peak in food availability for the birds occurring too soonfor the nestlings. As a further consequence, traditional population controls of O. brumata byP. major may become decoupled and result in unpredictable consequences for this commu-nity (Penuelas and Filella, 2001). Complex changes in herbivore and pathogen interactionswith plants, and their management, may also occur (Hepper, 2003; Badeck et al., 2004).

Springer

252 Urban Ecosyst (2006) 9:243–257

Walther et al. (2002) also discussed consequences for long distance migratory birds relyingon plants for food during and after migration. If birds obtain nectar and pollen from plantswhich flower earlier, then there may be less food available and increased interspecific andintraspecific competition.

Community changes may cause further changes in ecosystem functioning. Resource useand cycling by plants may change due to changes in plant resource needs for reproduction.Moreover, urbanization may modify community structure and the timing of pollen and nectaravailability which may in turn cause further modification due to new interspecific interactionssuch as competition for resources like water, nutrients, and pollinators.

Generally, changes in interspecific plant interactions will be complex and somewhat un-predictable. For instance, competition for resources (biotic and abiotic) could increase ifmore plants are flowering earlier (Fitter and Fitter, 2002) because reproduction requires a lotof resources. Urbanization effects on flowering phenology may also result in unpredictablechanges in ecosystem structure because species previously able to coexist due to niche dif-ferentiation may interact differently (Fitter and Fitter, 2002). Another potential communityaffect involves pollination facilitation. This controversial theory states that flower display ofone plant may help attract pollinators to neighboring plants, facilitating greater pollinationthan would occur if the plant were alone—even at high densities. The pollination facilitationtheory was introduced since plants are known to facilitate the growth of other plant speciesvia microclimate amelioration (Rathcke and Lacey, 1985; Feldman et al., 2004). Only a fewexamples of this phenomenon are known, contributing to the controversy (Feldman et al.,2004). If pollination facilitation does occur and is important for some species, then changesin flowering phenology may be beneficial or detrimental depending on the specific speciesinvolved. Non-indigenous flora and fauna in urban centers add another level of complexitywhose effects are not fully understood at this time. This area is ripe for further investigation—particularly with the abundance of non-indigenous plants prevalent in urban ecosystems.

Human health

The impact of changes in flowering phenology on human health has focused on pollenallergies, or pollinosis. Advancement in allergy seasons and their peaks (by 3–22 days),lengthening of allergy seasons, and increase in the severity of allergies have been reportedworld-wide (Van Vliet et al., 2002; Traidl-Hoffmann et al., 2003), and correlated with highertemperatures and possibly higher CO2 concentrations. Van Vliet et al. (2002) also found thatgrasses have advanced flowering by fewer days than many trees and shrubs. This observationfits with other studies that report less of an effect on wind-pollinated plants. On the otherhand, this seems contradictory to reports of earlier allergy seasons since the most widespreadknown offenders are wind-pollinated grass, ragweed, tree, and shrub species. The work byZiska et al. (2003) is apparently one of the few in situ experiments to study allergenicity andphenology of an allergenic plant along an urban-rural gradient. More studies are needed tobetter understand how changes in flowering phenology are related to the different aspects ofhuman allergies.

Economic ramifications

Economic impacts of changes in flowering phenology within urban centers have not beenstudied although those of general phenological changes have to some extent. A few studiesexamining global climate change and urban heat island effects on agriculture and other plant-based food production suggest important implications for agriculture and plant-based food

Springer

Urban Ecosyst (2006) 9:243–257 253

production within and/or close to urban centers. For instance, Baker et al. (2002) reported thatin the Phoenix metropolitan area from 1997–2000 cotton was planted about 14 days earlierin agriculture fields within the city and city fringe areas than in agriculture fields outside thereach of the urban heat island footprint. This implies that plants close to the city floweredearlier than those outside of the city. Moreover, even though cotton can self-pollinate, morecotton is obtained when pollinators are present (Rhodes, 2002). The advancement in cottonplanting may have implications for pollination by insects—especially bees—if the pollinatinginsects are not (as) available when the plants are blooming. If pollinators are not available,then there may be less yield. On the other hand, if the pollinating insects are developingquicker due to increased heat exposure, then there may not be significant differences. Inaddition, changes in flowering phenology may impact the quality and quantity of chemicalsubstance extracted from plants (Badeck et al., 2004).

Liu et al. (2004) hypothesized that a reduction in honey production in China in the 1990smay have resulted from warmer temperatures during that decade. Cape (2003) also discussesthe potential impacts of VOCs on agriculture—specifically on plants grown for their flow-ers. The economic costs of increased allergy problems should also be considered. Finally,changing flowering phenology could also have implications for gardeners and landscapers.Hepper (2003) reported earlier flowering of common garden perennials in Britain. Successfulplantings from seeds or from started stock from stores and planned landscaping effects (fromflower displays) may change due to urbanization effects on microclimate.

Discussion and conclusions

Several important patterns have emerged from this review of urbanization and floweringphenology (Table 1 and Fig. 1). First, there is a general trend towards earlier spring-timeflowering in urbanized areas compared to surrounding rural areas; although, many plantsalso show delayed flowering or no significant change in flowering. Second, some types ofplants appear to be affected more strongly by urbanization than others; specifically, insect-pollinated plants, early-spring bloomers, annuals, and non-woody plants. Third, temperatureappears to be the most frequently cited cause for changes in flowering phenology.

From this review, we can identify several knowledge gaps that need to be addressed by thefollowing research questions. First, will we see earlier flowering in urbanized areas in tropical,arid, or semi-arid ecosystems? The existing studies have focused only on temperate forest,Mediterranean, and boreal ecosystems which all have a single distinct growing season. Willplants in urban areas in aseasonal tropical forest ecosystems, desert ecosystems, or seasonsdelineated by moisture (e.g., wet/dry) show similar phenological affects?

Second, which kind of plants do not show changes phenological changes related to urban-ization and why? Is it because of photoperiod or interactions between phenological triggersthat “cancel” changes in each other? How will plants dominated by moisture or moisture-temperature phenological triggers be affected by urbanization? Does ecological light pollu-tion affect plants dominated by photoperiod triggers?

Third, what is the fine-scale spatial pattern of flowering phenology in cities? Does thespatiotemporal pattern of flowering change within a city? Are changes in the timing of plantflowering consistent along an urban-rural gradient? Or, will plants show spatial or temporalpatchiness in flowering across the urban landscape within a city? Researchers have studiedflowering patterns in rural ecosystems to decipher resource use by plants, resource use byanimals, ecosystem structure, niche differentiation, and evolution (e.g., Gentry, 1974; Pavonand Briones, 2001; Lobo et al., 2003). This work should also be done for different urban

Springer

254 Urban Ecosyst (2006) 9:243–257

Fig. 1 Diagram summarizing urbanization factor that can affect phenological triggers and the resultingpotential effect. Dashed lines indicate weak or controversial triggers of flowering. ∗ Method of study

ecosystems. Moreover, these kinds of studies will reveal more about how socioeconomicfactors, such as land cover type and land use affect flowering phenology and, therefore, inter-specific and intraspecific interactions. For instance, are plants of the same species becomingmore or less reproductively isolated? Is there more or less competition for resources for repro-duction? Is there higher phenotypic plasticity in flowering phenology in urban ecosystems?

Lastly, how do non-indigenous species affect urban ecosystem structure and function?Non-indigenous species (flora and fauna) are a fact for urban ecosystems. Are non-indigenousplants flowering at the same times as the indigenous plants? Do they provide pollen and nectarfor indigenous animals, non-indigenous animals, or both? Do they out-compete indigenousplants by taking pollinators away? Are non-indigenous plants, or cultivars, hybridizing withindigenous plants?

Springer

Urban Ecosyst (2006) 9:243–257 255

Understanding the causes, mechanisms, and consequences of changes in flowering phenol-ogy in urban environment has a number of important ecological and economic implications. Itis time that we move beyond descriptive studies that show differences in phenology betweenurban areas and surrounding rural environments. We now need to focus more on mechanis-tic studies that address what, how, and why plants are affected and what are the ecologicaland environmental consequences of these effects. To achieve this goal, a spatially explicitframework based on landscape ecology (Turner et al., 2001; Wu and Hobbs, 2002) shouldbe used that integrates spatial patterns of flowering phenology and environmental factorswith various causes and consequences of phenological changes on multiple scales. While thepotential is yet to be explored for studying flowering phenology, such a landscape ecologicalapproach has been quite successful in a variety of ecological studies, including a number ofurban ecological examples (e.g., Pickett et al., 2001; Luck and Wu, 2002; Wu and David,2002). While remote sensing is helpful in many ways, most of these questions have to beaddressed by direct observation or experimentation in the field.

Acknowledgments This research is in part supported by National Science Foundation (NSF) under GrantNo. DEB 9714833, Central Arizona–Phoenix Long-Term Ecological Research (CAP LTER) and by NationalScience Foundation (NSF) Grant No. DGB 9987612, Integrative Graduate Education and Research Training inUrban Ecology (IGERT). We would also like to thank the anonymous reviewers for their valuable comments.

References

Abd El-Ghani MM (1997) Phenology of ten common plant species in western Saudi Arabia. J Arid Environm35:673–683

Adebayo YR (1991) Daytime effects of urbanization on relative-humidity and vapor-pressure in a tropicalcity. Theoret Appl Climat 43:17–30

Aizen MA, Ashworth L, Galetto L (2002) Reproductive success in fragmented habitats: Do compatibilitysystems and pollination specialization matter? J Veget Sci 13:885–892

Augspurger CK (1983) Phenology, flowering synchrony, and fruit-set of 6 neotropical shrubs. Biotropica15:257–267

Badeck FW, Bondeau A, Bottcher K, Doktor D, Lucht W, Schaber J, Sitch S (2004) Responses of springphenology to climate change. New Phytol 162:295–309

Baker LA, Brazel AJ, Selover N, Martin C, McIntyre N, Steiner FR, Nelson A, Musacchio L (2002) Urban-ization and warming of Phoenix (Arizona, USA): Impacts, feedbacks and mitigation. Urban Ecosystems6:183–203

Battey NH (2000) Aspects of seasonality. J Experim Botany 51:1769–1780Bishop JG, Schemske DW (1998) Variation in flowering phenology and its consequences for lupines colonizing

Mount St. Helens. Ecology 79:534–546Blankenship SM, Bailey DA, Miller JE (1993) Effects of continuous, low-levels of ethylene on growth and

flowering of easter lily. Scientia Horticulturae 53:311–317Borchert R, Rivera G (2001) Photoperiodic control of seasonal development and dormancy in tropical stem-

succulent trees. Tree Physiol 21:213–221Bowers JE, Dimmitt MA (1994) Flowering phenology of 6 woody-plants in the Northern Sonoran Desert.

Bull Torrey Botan Club 121:215–229Brody AK (1997) Effects of pollinators, herbivores, and seed predators on flowering phenology. Ecology

78:1624–1631Cape JN (2003) Effects of airborne volatile organic compounds on plants. Environmental Pollution 122:145–

157Chauhan SVS, Chaurasia L, Rana A (2004) Impact of air pollution on floral morphology of Cassia siamea

Lamk. J Environm Bio 25:291–297Chow SD (1992) The urban climate of Shanghai. Atmosp Environm Part B-Urban Atmosp 26:9–15Chuine I, Cambon G, Comtois P (2000) Scaling phenology from the local to the regional level: Advances

from species-specific phenological models. Global Change Bio 6:943–952Conner JK, Neumeier R (2002) The effects of ultraviolet-B radiation and intraspecific competition on growth,

pollination success, and lifetime female fitness in Phacelia campanularia and P-purshii (Hydrophyllaceae).Amer J Botany 89:103–110

Springer

256 Urban Ecosyst (2006) 9:243–257

Crawford DL (2001). Light pollution: Changing the situation to everyone’s advantage. In: Preserving theAstronomical Sky, pp. 33–38. ASTRONOMICAL SOC PACIFIC, San Francisco

Day TA, Gober P, Xiong FSS, Wentz EA (2002) Temporal patterns in near-surface CO2 concentrations overcontrasting vegetation types in the Phoenix metropolitan area. Agricult Forest Meteor 110:229–245

Deen W, Hunt T, Swanton CJ (1998) Influence of temperature, photoperiod, and irradiance on the phenologicaldevelopment of common ragweed (Ambrosia artemisiifolia). Weed Sci 46:555–560

Feldman TS, Morris WF, Wilson WG (2004) When can two plant species facilitate each other’s pollination?Oikos 105:197–207

Fischer M, van Kleunen M, Schmid B (2000) Genetic Allee effects on performance, plasticity and develop-mental stability in a clonal plant. Ecol Lett 3:530–539

Fitter AH, Fitter RSR (2002) Rapid changes in flowering time in British plants. Science 296:1689–1691Franklin KA, Whitelam GC (2004) Light signals, phytochromes and cross-talk with other environmental cues.

J Experim Botany 55:271–276Friedel MH, Nelson DJ, Sparrow AD, Kinloch JE, Maconochie JR (1993) What induces central Australian

arid zone trees and shrubs to flower and fruit. Austral J Botany 41:307–319Gentry AH (1974) Coevolutionary patterns in central American bignoniaceae. Ann Miss Bota Gard 61:728–

759Hendry AP, Day T (2005) Population structure attributable to reproductive time: Isolation by time and adap-

tation by time. Molec Ecol 14:901–916Hepper FN (2003) Phenological records of English garden plants in Leeds (Yorkshire) and Richmond (Surrey)

from 1946 to 2002. An analysis relating to global warming. Biodiv Conserv 12:2503–2520Idso CD, Idso SB, Balling RC (2001) An intensive two-week study of an urban CO2 dome in Phoenix, Arizona,

USA. Atmosph Environm 35:995–1000Jonsson P (2004) Vegetation as an urban climate control in the subtropical city of Gaborone, Botswana. Intern

J Climat 24:1307–1322Kakani VG, Reddy KR, Zhao D, Sailaja K (2003) Field crop responses to ultraviolet-B radiation: A review.

Agricult Forest Meteorol 120:191–218Kang SY, Running SW, Lim JH, Zhao MS, Park CR, Loehman R (2003) A regional phenology model for

detecting onset of greenness in temperate mixed forests, Korea: An application of MODIS leaf area index.Remote Sens Environm 86:232–242

Kudo G, Nishikawa Y, Kasagi T, Kosuge S (2004) Does seed production of spring ephemerals decrease whenspring comes early? Ecolog Res 19:255–259

Lipfert FW, Cohen S, Dupuis LR, Peters J (1991) Relative-humidity predictor equations based onenvironmental-factors. Atmosp Environm Part B-Urban Atmosp 25:435–441

Liu FL, He JZ, Zhang XW, Fu WJ (2004) Lower Chinese honey production blamed on flowering delay. AmerBee J 144:142–142

Lobo JA, Quesada M, Stoner KE, Fuchs EJ, Herrerias-Diego Y, Rojas J, Saborio G (2003) Factors affectingphenological patterns of bombacaceous trees in seasonal forests in Costa Rica and Mexico. Amer J Botany90:1054–1063

Longcore T, Rich C (2004) Ecological light pollution. Front Ecol Environm 2:191–198Luck M, Wu JG (2002) A gradient analysis of urban landscape pattern: A case study from the Phoenix

metropolitan region, Arizona, USA. Lands Ecol 17:327–339Murren CJ (2002) Effects of habitat fragmentation on pollination: Pollinators, pollinia viability and reproduc-

tive success. J Ecol 90:100–107Mussey GJ, Potter DA (1997) Phenological correlations between flowering plants and activity of urban land-

scape pests in Kentucky. J Econ Entomol 90:1615–1627Nasrallah HA, Balling RC, Madi SM, Al-Ansari L (2003) Temporal variations in atmospheric CO2 concentra-

tions in Kuwait City, Kuwait with comparisons to Phoenix, Arizona, USA. Environm Poll 121:301–305Osman AII, Isobe S, Nawar S, Morcos AB (2001) Light pollution and energy loss from Cairo. In: Preserving

the Astronomical Sky, ASTRONOMICAL SOC PACIFIC, San Francisco, pp 107–110Papayannis A, Balis D, Bais A, Van Der Bergh H, Calpini B, Durieux E, Fiorani L, Jaquet L, Ziomas I,

Zerefos CS (1998) Role of urban and suburban aerosols on solar UV radiation over Athens, Greece.Atmosp Environm 32:2193–2201

Pauley SM (2004) Lighting for the human circadian clock: Recent research indicates that lighting has becomea public health issue. Medical Hypotheses 63:588–596

Pavon NP, Briones O (2001) Phenological patterns of nine perennial plants in an intertropical semi-aridMexican scrub. J Arid Environm 49:265–277

Penuelas J, Filella I (2001) Phenology—Responses to a warming world. Sci 294:793–794Penuelas J, Filella I, Zhang XY, Llorens L, Ogaya R, Lloret F, Comas P, Estiarte M, Terradas J (2004) Complex

spatiotemporal phenological shifts as a response to rainfall changes. New Phytolo 161:837–846

Springer

Urban Ecosyst (2006) 9:243–257 257

Petit S (2001) The reproductive phenology of three sympatric species of columnar cacti on Curacao. J AridEnvironm 49:521–531

Pickett STA, Cadenasso ML, Grove JM, Nilon CH, Pouyat RV, Zipperer WC, Costanza R (2001) Urban eco-logical systems: Linking terrestrial ecological, physical, and socioeconomic components of metropolitanareas. Ann Rev Ecol System 32:127–157

Primack D, Imbres C, Primack RB, Miller-Rushing AJ, Del Tredici P (2004) Herbarium specimens demonstrateearlier flowering times in response to warming in Boston. Amer J Botany 91:1260–1264

Rathcke B, Lacey EP (1985) Phenological patterns of terrestrial plants. Ann Rev Ecol System 16:179–214Repapis CC, Mantis HT, Paliatsos AG, Philandras CM, Bais AF, Meleti C (1998) Case study of UV-B

modification during episodes of urban air pollution. Atmosp Environm 32:2203–2208Rhodes J (2002) Cotton pollination by honey bees. Austral J Experim Agricult 42:513–518Rivera G, Borchert R (2001) Induction of flowering in tropical trees by a 30-min reduction in photoperiod:

Evidence from field observations and herbarium specimens. Tree Physiol 21:201–212Roetzer T, Wittenzeller M, Haeckel H, Nekovar J (2000) Phenology in central Europe—differences and trends

of spring phenophases in urban and rural areas. Intern J Biometeorol 44:60–66Rusterholz HP, Erhardt A (1998) Effects of elevated CO2 on flowering phenology and nectar production of

nectar plants important for butterflies of calcareous grasslands. Oecologia 113:341–349Sampson BJ, Cane JH (1999) Impact of enhanced ultraviolet-B radiation on flower, pollen, and nectar pro-

duction. Amer J Botany 86:108–114Santandreu M, Lloret F (1999) Effect of flowering phenology and habitat on pollen limitation in Erica multi-

flora. Canadian J Botany-Revue Canad De Botan 77:734–743Schwartz MD (2003) Phenology: An integrative environmental science. Kluwer, DordrechtSchwartz MD, Chen ZQ (2002) Examining the onset of spring in China. Clim Res 21:157–164Schwartz MD, Reed BC, White MA (2002) Assessing satellite-derived start-of-season measures in the con-

terminous USA. Intern J Climats 22:1793–1805Schwartz MD, Reiter BE (2000) Changes in North American spring. Intern J Climatol 20:929–932Sparks TH, Menzel A (2002) Observed changes in seasons: An overview. Intern J Climatol 22:1715–1725Traidl-Hoffmann C, Kasche A, Menzel A, Jakob T, Thiel M, Ring J, Behrendt H (2003) Impact of pollen on

human health: More than allergen carriers? Intern Arch Allergy Immunol 131:1–13Turner MG, Gardner RH, O’Neill RV (2001) Landscape ecology in theory and practice: Pattern and process.

Springer, New YorkUnger J (1999) Urban-rural air humidity differences in Szeged, Hungary. Intern J Climatol 19:1509–1515Van Vliet AJH, Overeem A, De Groot RS, Jacobs AFG, Spieksma FTM (2002) The influence of temperature

and climate change on the timing of pollen release in the Netherlands. Intern J Climatol 22:1757–1767Walther GR, Post E, Convey P, Menzel A, Parmesan C, Beebee TJC, Fromentin JM, Hoegh-Guldberg O,

Bairlein F (2002) Ecological responses to recent climate change. Nature 416:389–395Wcislo WT, Cane JH (1996) Floral resource utilization by solitary bees (Hymenoptera: Apoidea) and exploita-

tion of their stored foods by natural enemies. Ann Rev Entomol 41:257–286White MA, Nemani RR, Thornton PE, Running SW (2002) Satellite evidence of phenological differences

between urbanized and rural areas of the eastern United States deciduous broadleaf forest. Ecosystems5:260–273

Williams TA, Abberton MT (2004) Earlier flowering between 1962 and 2002 in agricultural varieties of whiteclover. Oecologia 138:122–126

Wu JG, David JL (2002) A spatially explicit hierarchical approach to modeling complex ecological systems:Theory and applications. Ecolog Modell 153:7–26

Wu JG, Hobbs R (2002) Key issues and research priorities in landscape ecology: An idiosyncratic synthesis.Lands Ecol 17:355–365

Yan WK, Wallace DH (1998) Simulation and prediction of plant phenology for five crops based on photoperiodx temperature interaction. Ann Botany 81:705–716

Yang TW, Abe Y (1974) Phenology of larrea-divaricata in tucson region. Amer J Botany 61:69–69Yanovsky MJ, Kay SA (2003) Living by the calendar: How plants know when to flower. Nat Rev Molec Cell

Bio 4:265–276Zhang XY, Friedl MA, Schaaf CB, Strahler AH (2004) Climate controls on vegetation phenological patterns

in northern mid- and high latitudes inferred from MODIS data. Global Change Bio 10:1133–1145Zhao TT, Schwartz MD (2003) Examining the onset of spring in Wisconsin. Clim Res 24:59–70Ziska LH, Gebhard DE, Frenz DA, Faulkner S, Singer BD, Straka J (2003) Cities as harbingers of climate

change: Common ragweed, urbanization, and public health. J Allergy Clini Immun 111:290–295

Springer